U.S. patent number 6,907,130 [Application Number 09/023,279] was granted by the patent office on 2005-06-14 for speech processing system and method using pseudospontaneous stimulation.
This patent grant is currently assigned to Research Triangle Institute, University of Iowa Research Foundation. Invention is credited to Jay Rubinstein, Blake Wilson.
United States Patent |
6,907,130 |
Rubinstein , et al. |
June 14, 2005 |
Speech processing system and method using pseudospontaneous
stimulation
Abstract
A apparatus and method for inner ear implants is provided that
generates signal processing stochastic independence activity across
the excited neural population. A high rate pulse train can produce
random spike patterns in auditory nerve fibers (hereafter
"pseudospontaneous activity") that are statistically similar to
those produced by spontaneous activity in the normal auditory
nerve. We call this activity "pseudospontaneous". Varying rates of
pseudospontaneous activity can be created by varying the intensity
of a fixed amplitude, high rate pulse train stimulus, e.g., 5000
pps. The high rate pulse train can desynchronize the nerve fiber
population and can be combined with a data signal in an inner ear
implant. The pseudospontaneous activity can enhance neural
representation of temporal detail and dynamic range with an inner
ear implant such as a cochlear implant. The pseudospontaneous
activity can further eliminate a major difference between
acoustic-and electrical-derived hearing percepts.
Inventors: |
Rubinstein; Jay (Solon, IA),
Wilson; Blake (Durham, NC) |
Assignee: |
University of Iowa Research
Foundation (Iowa City, IA)
Research Triangle Institute (Research Triangle Park,
NC)
|
Family
ID: |
21814150 |
Appl.
No.: |
09/023,279 |
Filed: |
February 13, 1998 |
Current U.S.
Class: |
381/312; 381/151;
600/25; 607/56; 623/10; 607/57; 607/55; 381/331; 381/326 |
Current CPC
Class: |
A61N
1/36038 (20170801) |
Current International
Class: |
A61N
1/36 (20060101); H04R 25/00 (20060101); A61F
2/18 (20060101); H04R 025/00 () |
Field of
Search: |
;607/55-57,137 ;600/25
;381/326,331,151,312,320,316,FOR 130/ ;623/10 ;455/40,41,43 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
2 171 605 |
|
Sep 1986 |
|
GB |
|
WO9612383 |
|
Apr 1996 |
|
WO |
|
Other References
Ifukube et al., "Design Of An Implantable Tinnitus Suppressor By
Electrical Cochlear Stimulation", Biomechanics, Rehabilitation,
Electrical Phenomena, Biomaterials, San Diego, Oct. 28-31, 1993,
vol. 3, No. CONF. 15, pp. 1349-1350. .
Cohen, N.L. et al., "A Prospective, Randomized Study of Cochlear
Implants," N. Engl. J. Med., 328:233-7, 1993. .
C.W. Parkins et al., "A Fiber Sum Modulation Code for a Cochlear
Prosthesis" Annals of the New York Academy of Sciences, Jan. 1,
1983, vol. 405, pp. 490-501. .
P.C. Loizou, "Signal Processing for Cochlear Prosthesis: A Tutorial
Review", Proceedings of the 40th Midwest Symposium on Circuits and
Systems MWSCAS, IEEE 1997, pp. 881-885..
|
Primary Examiner: Kuntz; Curtis
Assistant Examiner: Harvey; Dionne
Attorney, Agent or Firm: Fleshner & Kim, LLP
Government Interests
Part of the work performed during the development of this invention
utilized U.S. Government funds under grant DC 6211 and contract OD
02948 from the National Institute of Health. The government may
have certain rights in this invention.
Claims
What is claimed is:
1. A cochlear system, comprising: a signal generator that generates
a second signal capable of causing pseudospontaneous activity in an
auditory nerve; a signal processor that combines a first signal
that represents sound and the second signal to output a combined
signal; and a stimulation unit coupled to the signal processor that
receives the combined signal from the signal processor, wherein the
stimulation unit is configured to apply the combined signal to the
auditory nerve.
2. The system according to claim 1, wherein the stimulation unit is
an electrode array unit that is coupled to the auditory nerve.
3. The system according to claim 2, wherein the first signal is
applied to a first subset of electrodes in the electrode array and
the second signal is applied to a second subset of electrodes in
the electrode array.
4. The system according to claim 1, wherein the second signal
includes one of (i) a pulse train generating substantially
continuous pseudospontaneous activity, (ii) a broad band noise, and
(iii) at least fluctuations in amplitude greater than prescribed
amount at a frequency above approximately 2 k Hz that causes
statistically independent activity in a plurality of nerve fibers
of the nerve.
5. The system according to claim 1, wherein the pseudospontaneous
activity is demonstrated by statistically independent activity in a
plurality of nerve fibers in the auditory nerve.
6. The system according to claim 1, wherein the second signal
includes rapid state transitions and a frequency greater than
approximately 3 kilohertz.
7. The system according to claim 1, wherein the signal processor
determines the combined signal by summing the first and second
signals.
8. The system according to claim 1, further comprising a microphone
that generates the first signal, wherein the microphone is coupled
to the signal processor.
9. The system according to claim 1, wherein the signal processor
further comprises a combining circuit that logically processes the
first and second signals, wherein the combining circuit ANDs the
first and second signals.
10. The system according to claim 1, wherein a microphone, the
signal processor and the signal generator are positioned external
to an ear, wherein the stimulation unit is coupled by a wire to the
signal processor, and wherein the stimulation unit is coupled to
the auditory nerve via a cochlea.
11. A method for generating a driving signal for an auditory
implant, comprising: receiving a first signal; generating a second
signal that causes pseudo-spontaneous activity in an acoustic
nerve; combining the first and second signals to generate the
driving signal; and applying the combined signal to the acoustic
nerve.
12. The method according to claim 11, wherein the first signal
represents at least one of speech, emergency signals and control
information.
13. The method according to claim 11, wherein the combining step
performs at least one of summing and multiplying the first and
second signals.
14. The method of claim 11, wherein the applying the combined
signal generates substantially continuous pseudospontaneous
activity.
15. The method of claim 11, wherein the second signal is not
continuously applied.
16. The method of claim 11, wherein the second signal includes one
of (i) a pulse train generating substantially continuous
pseudospontaneous activity, (ii) a broad band noise, and (iii) at
least fluctuations in amplitude greater than prescribed amount at a
frequency above approximately 2 k Hz that causes statistically
independent activity in a plurality of nerve fibers of the nerve,
wherein the driving signal is used to modulate a carrier
signal.
17. An auditory prosthesis for receiving an auditory signal
representing sound and supplying an electrical signal which is
adapted to stimulate the auditory nerve of a person, comprising:
pseudospontaneous generation means for generating a
pseudospontaneous driving signal; transducer means adapted to
receive the auditory signal and the pseudospontaneous driving
signal for transforming the auditory signal and the
pseudospontaneous driving signal to electrical input signals; and
stimulation means, operatively coupled to the electrical input
signals generated by the transducer means, for stimulating the
auditory nerve at defined locations within the cochlea, wherein at
least one of the plurality of electrical signals is configured to
cause statistically independent activity in a plurality of nerve
fibers of an auditory nerve.
18. The auditory prosthesis of claim 17, wherein the transducer
means further performs at least one of the summing and multiplying
the auditory signal and the pseudo-spontaneous driving signal.
19. The auditory prosthesis of claim 17, wherein the
pseudospontaneous driving signal includes one of (i) a pulse train
generating substantially continuous activation, (ii) a broad band
noise, or (iii) at least fluctuations in amplitude greater than
prescribed amount at a frequency above approximately 2 k Hz,
wherein the electrical signals stimulate the auditory nerve.
20. A neural prosthetic apparatus, comprising. a signal generator
that generates a second signal; a signal processor that combines a
first signal that represents sound and the second signal to output
a combined signal, wherein a carrier signal is modulated with the
combined signal; and stimulation unit coupled to the signal
processor that receives and demodulates the carrier signal to
obtain the combined signal from the signal processor for
application to the auditory nerve, wherein the second signal
includes at least fluctuations in amplitude greater than a
prescribed amount at a frequency above approximately 2 kHz.
21. The apparatus according to claim 20, wherein the stimulation
unit is an electrode array unit that is coupled to the auditory
nerve, and wherein the first signal is applied to a first subset of
electrodes in the electrode array and the second signal is applied
to a second subset of electrodes in the electrode array.
22. The apparatus according to claim 20, wherein the second signal
generates statistically independent activity in a plurality of
nerve fibers in the auditory nerve.
23. The apparatus according to claim 20, wherein the auditory nerve
comprises a plurality of nerve fibers, and wherein the second
signal comprises one or more signals that generate a substantially
maximum firing rate of the plurality of nerve fibers.
24. The apparatus according to claim 20, wherein the second signal
includes one of (i) a pulse train generating substantially
continuous pseudospontaneous activity being statistically
independent activity in a plurality of nerve fibers of the nerve,
and (ii) a broad band noise that causes statistically independent
activity in the plurality of nerve fibers of the nerve.
25. The apparatus according to claim 20, wherein the prosthesis is
a cochlear implant applying current to the auditory nerve, wherein
the stimulation unit is configured to apply the combined signal to
the auditory nerve.
26. The apparatus according to claim 20, wherein the
pseudospontaneous activity continues after the second signal is
stopped.
27. A method of modifying a neural prosthetic apparatus that
receives an information signal and supplies a corresponding
electrical signal to stimulate an auditory nerve, comprising:
providing a pseudospontaneous signal generator that generates a
second signal; and providing an electrical coupling means for
supporting an electrical connection from the pseudospontaneous
signal generator to at least one electrical contact, and wherein
the second signal is configured to induce a random pattern of
activation in the auditory nerve mimicking the spontaneous neural
activity of the auditory nerve.
28. The method of claim 27, wherein the information signal
represents at least one of speech, emergency signals and control
information, and wherein the second signal includes one of (i) a
pulse train generating substantially continuous pseudospontaneous
activity, (ii) a broad band noise, (iii) at least fluctuations in
amplitude greater than prescribed amount at a frequency above
approximately 2 k Hz, and (iv) at least fluctuations in amplitude
greater than prescribed amount at a frequency that causes
statistically independent activity in a plurality of nerve fibers
of the auditory nerve.
29. The method of claim 27, wherein the neural prosthetic apparatus
is a cochlear implant, wherein the second signal and the electrical
signal are used to modulate a carrier signal.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to an system and method for
providing pseudospontaneous neural stimulation. In particular, the
invention relates to an apparatus and method for providing
pseudospontaneous activity in the auditory nerve, which can be used
to treat a sensorineural deafness patient. Electrical signals that
induce pseudospontaneous neural activity in the auditory nerve can
be delivered to the patient via an inner ear (cochlear)
implant.
2. Background of the Related Art
At least two distinct types of hearing problems are recognized:
conductive hearing loss and sensorineural hearing loss. The former
is generally due to a mechanical defect in the middle ear that
prevents sound-related vibrations from reaching the inner ear. In
the latter, sound-related vibrations reach the inner ear, but
signal transmission to the brain does not occur or is restricted.
Sensorineural hearing loss usually results from damage to the
cochlea and/or the auditory nerve. Sensorineural hearing loss is a
common condition that may occur in old age, or may be due to
exposure to excessively loud noises (e.g. rock concerts, jet
engines), viral infections, etc.
Patients experiencing a certain amount of hearing loss may benefit
from the use of a hearing aid which increases the volume of sound
electronically, and which may be placed either behind the pinna of
the ear or within the outer ear canal. In both cases, the device
usually comprises a microphone for transforming sound waves into
electrical signals, an amplifier for increasing the strength of the
electrical signals, and an earphone for providing amplified sounds.
Devices designed to treat deafness must obviously consider the
underlying cause of deafness. For example, a sensorineural deafness
patient with a defective cochlea who still has a functional
auditory nerve, may benefit from a cochlear implant, as described
hereinbelow. However, if the auditory nerve is itself damaged and
cannot carry electrical signals, then the problem is "too far
downstream" in the signal processing sequence for a cochlear
implant to be effective. In that situation, artificial signals must
enter the auditory system "beyond the block" for example, in the
brain stem or in the auditory cortex.
Cochlear implants were designed for patients who are deaf as a
result of loss of the cochlea's sound transduction mechanism. In
this situation, an electrode is implanted in the cochlea whereby
the electrode, upon receiving electrical signals from a speech
processor directly stimulates the auditory nerve. Consequently,
candidates for a cochlear implant device must have an intact
auditory nerve capable of carrying electrical signals to the brain
stem. The cochlear implant device delivers electrical signals e.g.,
by means of a multi-contact stimulating electrode. The stimulating
electrode is surgically inserted by an otolaryngologist into the
damaged cochlea. Activation of the contacts stimulates auditory
nerve terminals that are normally activated by the cochlear sound
transduction mechanism (hair cells-spiral ganglion). The patient
perceives sound as the coded electrical signal carried into the
brain by the auditory nerve. (See for example, Cohen, N. L. et al.,
"A Prospective, Randomized Study of Cochlear Implants," N. Engl. J.
Med., 328:233-7, 1993.)
Cochlear implants are surgically placed in the cochlea within the
temporal bone with little risk to the patient, because patients who
are already deaf due to a defective cochlea have little chance of
any additional injury being caused by placement of a cochlear
implant. In patients with hearing loss caused by dysfunction at the
level of the cochlea, cochlear implants can restore hearing.
However, fundamental differences currently exist between electrical
stimulation and acoustic stimulation of the auditory nerve.
Electrical stimulation of the auditory nerve, for example, via a
cochlear implant, generally results in more cross-fiber synchrony,
less within fiber jitter, and less dynamic range, as compared with
acoustic stimulation which occurs in individuals having normal
hearing. As a result, hearing percepts experienced by sensorineural
deafness patients via a cochlear implant lack the coherence and
clarity characteristics of normal hearing.
FIG. 15 shows a related art pattern of electrically-evoked compound
action potentials (EAPs) magnitudes from an auditory nerve of a
human subject with an electrical stimulus of 1 kHz (1016 pulses/s).
The EAP magnitudes are normalized to the magnitude of the first EAP
in the record. FIG. 15 shows the typical alternating pattern
previously described in the art. This pattern arises because of the
refractory period of the nerve and can degrade the neural
representation of the stimulus envelope. With a first stimulus
1502, a large response occurs likely because of synchronous
activation of a large number of nerve fibers. These fibers are
subsequently refractory during a second pulse 1504, and
accordingly, a small response is generated. By the time of a third
pulse 1506, an increased pool of fibers becomes available and the
corresponding response increases. The alternating synchronized
response pattern can be caused by a lack or decrease of spontaneous
activity in the auditory nerve and can continue indefinitely.
The above reference is incorporated by reference herein where
appropriate for appropriate teachings of additional or alternative
details, features and/or technical background.
SUMMARY OF THE INVENTION
An object of the present invention is to provide an apparatus and
method for neural stimulation that substantially obviates at least
the problems and disadvantages of the related art.
Another object of the present invention is to provide an apparatus
and method that generates stochastically independent or
pseudospontaneous neural activity.
A further object of the present invention is to provide an
apparatus and method that generates pseudospontaneous activity in
an auditory nerve to improve response to signals representing
sound.
A further object of the present invention is to provide an
apparatus and method that combines a conditioner signal and a data
signal and provides the combined signals to a neural system to
improve the response of the neural system to the data signal.
A further object of the present invention is to provide a cochlear
implant, and method for using same, that provides a conditioner and
a data signal to improve speech communications in a sensorineural
deafness patient via a cochlear implant.
To achieve at least the above objects in whole or in part, there is
provided a cochlear implant system according to the present
invention that includes a first signal that represents sound, a
signal generator that generates a second signal causing
pseudospontaneous activity, a signal processor that combines a
first signal and the second signal to output combined signals, and
a stimulation unit coupled to the signal processor that receives
the combined signal from the signal processor.
To further achieve at least the above objects in a whole or in
parts, there is provided a method for generating a driving signal
for an auditory implant according to the present invention that
includes receiving a first signal, generating a second signal that
causes pseudospontaneous activity in an auditory nerve and
combining the first and second signals to generate the driving
signal.
To further achieve at least the above objects in a whole or in
parts, there is provided an auditory prosthesis according to the
present invention for receiving an auditory signal representing
sound and supplying an electrical signal that is adapted to
stimulate the auditory nerve of a person that includes a
pseudospontaneous generation device that generates a
pseudospontaneous driving signal, a transducer device adapted to
receive the auditory signal and the pseudospontaneous driving
signal that transforms the signal to an electrical input signal, a
generation unit operatively coupled to the electrical input signal
that generates a plurality of electrical signals selectively
replicating the temporal nerve discharge pattern of individually
located auditory nerve fibers within the cochlea of a person and a
stimulation device, operatively coupled to the plurality of
electrical signals of the generation unit, that stimulates selected
auditory nerve sites within the cochlea.
Additional advantages, objects, and features of the invention will
be set forth in part in the description which follows and in part
will become apparent to those having ordinary skill in the art upon
examination of the following or may be learned from practice of the
invention. The objects and advantages of the invention may be
realized and attained as particularly pointed out in the appended
claims.
BRIEF DESCRIPTION OF THE DRAWINGS
The invention will be described in detail with reference to the
following drawings in which like reference numerals refer to like
elements wherein:
FIG. 1 is a diagram showing section view of the human ear as seen
from the front;
FIGS. 2A and 2B are diagrams showing the relative positions of the
hearing elements including the external ear, auditory cortex,
cochlea and cochlear nucleus;
FIG. 3A is a diagram showing neuronal membrane potential during
transmission of a nerve impulse;
FIG. 3B is a diagram showing the changes in permeability of the
plasma membrane to Na+ and K+ during the generation of an action
potential;
FIGS. 4A and 4B are diagrams showing histograms of modeled
responses of the human auditory nerve to a high rate pulse
train;
FIGS. 5A-5D are diagrams showing interval histograms of modeled
responses of the human auditory nerve to a high rate pulse train at
various intensities;
FIG. 6 is a diagram showing a relationship between stimulus
intensity and pseudospontaneous rate;
FIG. 7 is a diagram showing a relationship between stimulus
intensity and vector strength;
FIG. 8A is a diagram showing two exemplary unit waveforms;
FIG. 8B is a diagram showing an interval histogram;
FIGS. 8C-8D are diagrams showing exemplary recurrence time
data;
FIG. 9 is a diagram showing magnitudes of EAPs from a human subject
with variable rate pulse train;
FIG. 10 is a diagram showing magnitudes of EAPs produced with
stimulation of implant subject intracochlear electrode with various
combinations of conditioner and stimulus;
FIG. 11 is a diagram showing a preferred embodiment of a driving
signal for an auditory nerve according to the present
invention;
FIG. 12 is a diagram showing a preferred embodiment of an inner ear
stimulation system according to the present invention;
FIGS. 13A and 13B are diagrams showing exemplary implementations of
the inner ear stimulation system of FIG. 12;
FIG. 14 is a flowchart showing of a preferred embodiment of a
method for speech processing using pseudospontaneous stimulation of
the auditory nerve; and
FIG. 15 is a diagram showing related art EAP N1P1 magnitudes from a
human subjected to a stimulus.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
The auditory system is composed of many structural components, some
of which are connected extensively by bundles of nerve fibers. The
auditory system enables humans to extract usable information from
sounds in the environment. By transducing acoustic signals into
electrical signals, which are processed in the brain, humans can
discriminate among a wide range of sounds with great precision.
FIG. 1 shows a sectional view of a human ear 5, which includes the
outer ear 5A, middle ear 5B and inner ear 5C. The outer ear 5A
includes pinna 7 having folds of skin and cartilage and outer ear
canal 9, which leads from the pinna 7 at its proximal end to the
eardrum 11 at its distal end. The eardrum 11 includes a membrane
extending across the distal end of the outer ear canal 9. The
middle ear 5B is located between the eardrum 11 and the inner ear
5C and includes three small connected bones (ossicles), namely the
hammer 12, the anvil 14, and the stirrup 16. The hammer 12 is
connected to the inner portion of the eardrum 11, the stirrup 16 is
attached to oval window 20, and the anvil 14 is located between and
attached to each of the hammer 12 and the stirrup 16. A round or
oval window 20 leads to the inner ear 5C. The inner ear 5C includes
the labyrinth 27 and the cochlea 29, each of which is a
fluid-filled chamber. The labyrinth 27, which is involved in
balance, includes the semicircular canals 28. Vestibular nerve 31
attaches to the labyrinth 27. Cochlea 29 extends from the inner
side of the round window 20 in a generally spiral configuration,
and plays a key role in hearing by transducing vibrations
transmitted from middle ear 5B into electrical signals for
transmission along auditory nerve 33 to the hearing centers of the
brain (FIGS. 2A and 2B).
In normal hearing, sound waves collected by the pinna 7 are
funneled down the outer ear canal 9 and vibrate the eardrum 11. The
vibration is passed to the ossicles (hammer 12, anvil 14, and
stirrup 16). Vibrations pass through the round window 20 via the
stirrup 16 causing the fluid within the cochlea 29 to vibrate. The
cochlea 29 is equipped internally with a plurality of hair cells
(not shown). Neurotransmitters released by the hair cells stimulate
the auditory nerve 33 thereby initiating signal transmission along
the auditory nerve 33. In normal hearing, the inner hair
cell-spiral ganglion is inherently "noisy" in the absence of sound
because of the random release of neurotransmitters from hair cells.
Accordingly, in normal hearing, spontaneous activity in the
auditory nerve occurs in the absence of sound.
FIGS. 2A and 2B respectively show a side view and a front view of
areas involved in the hearing process, including the pinna 7 and
the cochlea 29. In particular, the normal transduction of sound
waves into electrical signals occurs in the cochlea 29 that is
located within the temporal bone (not shown). The cochlea 29 is
tonotopically organized, meaning different parts of the cochlea 29
respond optimally to different tones; one end of the cochlea 29
responds best to high frequency tones, while the other end responds
best to low frequency tones. The cochlea 29 converts the tones to
electrical signals that are then received by the cochlea nucleus
216, which is an important auditory structure located in the brain
stem 214. As the auditory nerve leaves the temporal bone and enters
the skull cavity, it penetrates the brain stem 214 and relays coded
signals to the cochlear nucleus 216, which is also tonotopically
organized. Through many fiber-tract interconnections and relays
(not shown), sound signals are analyzed at sites throughout the
brain stem 214 and the thalamus 220. The final signal analysis site
is the auditory cortex 222 situated in the temporal lobe 224.
Information is transmitted along neurons (nerve cells) via
electrical signals. In particular, sensory neurons such as those of
the auditory nerve carry information about sounds in the external
environment to the central nervous system (brain). Essentially all
cells maintain an electrical potential (i.e., the membrane
potential) across their membranes. However, nerve cells use
membrane potentials for the purpose of signal transmission between
different parts of an organism. In nerve cells, which are at rest
(i.e., not transmitting a nerve signal) the membrane potential is
referred to as the resting potential (Vm). The electrical
properties of the plasma membrane of nerve cells are subject to
abrupt change in response to a stimulus (e.g., from an electrical
impulse or the presence of neurotransmitter molecules), whereby the
resting potential undergoes a transient change called an action
potential. The action potential causes electrical signal
transmission along the axon (i.e., conductive core) of a nerve
cell. Steep gradients of both Na+ and K+ are maintained across the
plasma membranes of all cells via the Na--K pump.
TABLE 1 ION [INSIDE] (mM) [OUTSIDE] (mM) K+ 140 5 Na+ 10 145
Such gradients provide the energy required for both the resting
potential and the action potential of neurons. Concentration
gradients for Na+ and K+ (in the axon of a mammalian neuron) are
shown in Table 1. In a resting neuron, K+ is near electrochemical
equilibrium, while a large electrochemical gradient exists for Na+.
However, little trans-membrane movement of Na+ occurs because of
the relative impermeability of the membrane in the resting state.
In the resting state, the voltage-sensitive Na+ specific channels
and the voltage-sensitive K+ specific channels are both closed. The
passage of a nerve impulse along the axonal membrane is because of
a transient change in the permeability of the membrane, first to
Na+ and then to K+, which results in a predictable pattern of
electrical changes propagated along the membrane in the form of the
action potential.
The action potential of a neuron represents a transient
depolarization and repolarization of its membrane. As alluded to
above, the action potential is initiated by a stimulus, either from
a sensory cell (e.g., hair cell of the cochlea) or an electrical
impulse (e.g., an electrode of a cochlear implant). Specifically,
upon stimulation the membrane becomes locally depolarized because
of a rapid influx of Na+ through the voltage-sensitive Na+
channels. Current resulting from Na+ influx triggers depolarization
in an adjacent region of the membrane, whereby depolarization is
propagated along the axon. Following depolarization, the
voltage-sensitive K+ channels open. Hyperpolarization results
because of a rapid efflux of K+ ions, after which the membrane
returns to its resting state. (See, for example, W. M. Becker &
D. W. Deamer, The World of the Cell, 2nd Ed., pp. 616-640,
Benjamin/Cummings, 1991. (hereafter Becker)) The above sequence of
events requires only a few milliseconds.
FIG. 3A shows a membrane potential of a nerve cell during
elicitation of an action potential in response to a stimulus.
During generation of an action potential, the membrane first
becomes depolarized above a threshold level of at least 20 mV such
that the membrane is rendered transiently very permeable to Na+, as
shown in FIG. 3B, leading to a rapid influx of Na+. As a result,
the interior of the membrane becomes positive for an instant and
the membrane potential increases rapidly to about +40 mV. This
increased membrane potential causes an increase in the permeability
of the membrane to K+. A rapid efflux of K+ results and a negative
membrane potential is reestablished at a level below the resting
potential (Vm). In other words, the membrane becomes hyperpolarized
302 as shown in FIG. 3A. During this period of hyperpolarization
302, the sodium channels are inactivated and unable to respond to a
depolarization stimulus. The period 302 during which the sodium
channels, and therefore the axon, are unable to respond is called
the absolute refractory period. The absolute refractory period ends
when the membrane potential returns to the resting potential. At
resting potential, the nerve cell can again respond to a
depolarizing stimulus by the generation of an action potential. The
period for the entire response of a nerve cell to a depolarizing
stimulus, including the generation of an action potential and the
absolute refractory period, is about 2.5 to about 4 ms. (See, for
example, Becker, pp. 614-640)
As alluded to herein above, in a normal cochlea the inner hair
cell-spiral ganglion is inherently "noisy" (i.e., there is a high
background of activity in the absence of sound) resulting in
spontaneous activity in the auditory nerve. Further, sound produces
a slowly progressive response within and across fiber
synchronization as sound intensity is increased. The absence of
spontaneous activity in the auditory nerve can lead to tinnitus as
well as other hearing-related problems.
According to the preferred embodiments of the present invention,
the artificial induction of a random pattern of activation in the
auditory nerve of a tinnitus patient or a hard-of-hearing patient
mimics the spontaneous neural activation of the auditory nerve,
which routinely occurs in an individual with normal hearing and
lacking tinnitus. The artificially induced random pattern of
activation of the auditory nerve is hereafter called
"Pseudospontaneous". In the case of an individual having a damaged
cochlea, such induced pseudospontaneous stimulation activation of
the auditory nerve may be achieved, for example, by the delivery of
a high rate pulse train directly to the auditory nerve via a
cochlea implant. Alternatively, in the case of a patient with a
functional cochlea, pseudospontaneous stimulation of the auditory
nerve may be induced directly by stimulation via an appropriate
middle ear implantable device. Applicant has determined that by
inducing pseudospontaneous activity and desynchronizing the
auditory nerve, the symptoms of tinnitus may be alleviated.
Preferred embodiments of the present invention emphasize stochastic
independence across an excited neural population. A first preferred
embodiment of a neural driving signal according to the present
invention that generates pseudospontaneous neural activity will now
be described. In particular, high rate pulse trains according to
the first preferred embodiment can produce random spike patterns in
auditory nerve fibers that are statistically similar to those
produced by spontaneous activity in the normal spiral ganglion
cells. Simulations of a population of auditory nerve fibers
illustrate that varying rates of pseudospontaneous activity can be
created by varying the intensity of a fixed amplitude, high rate
pulse train stimulus. Further, electrically-evoked compound action
potentials (EAPs) recorded in a human cochlear implant subject
verify that such a stimulus can desynchronize the nerve fiber
population. Accordingly, the preferred embodiments according to the
present invention can eliminate a major difference between acoustic
and electric hearing. An exemplary high rate pulse train driving
signal 1102 according to the first embodiment is shown in FIG.
1.
A population of 300 modeled auditory nerve fibers (ANF) has been
simulated on a Cray C90 (vector processor) and IBM SP-2 (parallel
processors) system., The ANF model used a stochastic representation
of each node of Ranvier and a deterministic representation of the
internode. Recordings were simulated at the 13th node of Ranvier,
which approximately corresponds to the location of the porus of the
internal auditory canal assuming the peripheral process has
degenerated. Post-stimulus time (PST) histograms and interval
histograms were constructed using 10 ms binning of the peak of the
action potential. As is well-known in the art, a magnitude of the
EAPs is measured by the absolute difference in a negative peak (N1)
after pulse onsets and a positive peak (P2) after pulse onsets.
Stimuli presented to the ANF model were a high rate pulse train of
50 .mu.s monophasic pulses presented at 5 kHz for 18 ms from a
point source monopolar electrode located 500 .mu.m perpendicularly
from the peripheral terminals of the axon population. All acoustic
nerve fibers were simulated as being in the same geometric
location. Thus, each simulation can be considered to represent
either 300 fibers undergoing one stimulus presentation or a single
fiber undergoing 300 stimulus presentations. In addition, a first
stimulus of the pulse train was of sufficient magnitude to evoke a
highly synchronous spike in all 300 axons; all subsequent pulses
are of an equal, smaller intensity. The first stimulus
substantially increased computational efficiency by rendering all
fibers refractory with the first pulse of the pulse train.
Two fibers were simulated for eight seconds using the parameters
described above. Spike times were determined with one .mu.s
precision and assembled into 0.5 ms bins. Conditional mean
histograms, hazard functions and forward recurrence time histograms
were calculated (using 0.5 ms bins because of the small number of
spikes (1000) simulated) as known to one of ordinary skill in the
art. For example, see Analysis of Discharges Recorded
Simultaneously From Pairs of Auditory Nerve Fibers, D. H. Johnson
and N. Y. S. Kiang, Journal of Biophysics, 16, 1976, pages 719-734,
(hereafter Johnson and Kiang), hereby incorporated by reference.
See also "Pseudospontaneous Activity: Stochastic Independence of
Auditory Nerve Fibers with Electrical Stimulation," J. T.
Rubinstein, et al., pages 1-18, 1998, hereby incorporated by
reference.
FIG. 4A shows a post-stimulus time (PST) histogram 402 of discharge
times from the ANF model with a stimulus amplitude of 325 .mu.A. A
highly synchronous response 404 to a first, higher amplitude pulse
was followed by a "dead time" 406. Then, an increased probability
of firing 408 was followed by a fairly uniform firing probability
410. The y-axis of the PST histogram has been scaled to demonstrate
temporal details following the highly synchronous response to the
first pulse. There was a small degree of synchronization with the
stimulus as measured by a vector strength of 0.26.
FIG. 4B shows an interval histogram of the same spike train. As
shown in FIG. 4B, a dead time 412 was followed by a rapid increase
in probability 414 and then an exponential decay 416. The interval
histogram is consistent with a Poisson process following a dead
time, a renewal process, and greatly resembles interval histograms
of spontaneous activity in the intact auditory nerve. These
simulation results corresponds to a spontaneous rate of 116
spikes/second measured during the uniform response period of 7 to
17 ms.
As shown in FIGS. 5A-5D, when the stimulus intensity was varied in
the ANF model, the firing rate and shape of the PST and interval
histograms changed. FIGS. 5A-5D show four interval histograms of a
response to a 5 kHz pulse train at different stimulus intensities
that demonstrated a range of possible firing rates. The histograms
changed shape with changes in pseudospontaneous rate in a manner
consistent with normal auditory nerve fibers. All demonstrate
Poisson-type intervals following a dead-time. The firing rate
during the period of uniform response probability is given in the
upper right corner of each plot. Similarly, as respectively shown
in FIGS. 8 and 9, a conditional mean histogram and a hazard
function for a single "unit" simulated for eight seconds were
within standard deviations of theoretical limits. Thus, the
conditional mean histogram was "constant," which is consistent with
a renewal process, and indicated that a firing probability was not
affected by intervals prior to the previous spike. The hazard
function was also "constant" after a dead-time, followed by a
rapidly rising function. Thus, both plots were consistent with a
renewal process much like spontaneous activity, at least for the
intervals for which the ANF model had an adequate sample.
FIG. 6 shows the relationship between stimulus intensity and
pseudospontaneous rate. A full range of spontaneous rates,
previously known in animal, (from zero to approximately 150
spikes/s), was demonstrated over a relatively narrow range of
stimulus intensity for the high rate pulse train stimulation in a
computer simulation. Since there is minimal synchronization with
the stimulus, compound action potentials in response to individual
pulses would be expected to be small or unmeasurable.
Normal spontaneous activity is independent across neurons. Since
pseudospontaneous activity is driven by a common stimulus, one
measure of the relative degree of dependence/independence and
individual nerve fibers within the auditory nerve was vector
strength. Vector strength is a measure of the degree of periodicity
or synchrony with the stimulus. Vector strength is calculated from
period histograms and varies between 0 (no periodicity) and 1
(perfect periodicity). If vector strength is "high" then each fiber
will be tightly correlated with the stimulus and two such fibers
will be statistically dependent. If vector strength is "low" then
two such fibers should be independent. As shown in FIG. 7, a
relationship between stimulus intensity and vector strength is
nonzero, but is below or near a noise floor at all intensities
tested for the high rate pulse train stimulation. In addition,
there is little effect of stimulus amplitude on synchrony. A noise
floor for the vector strength calculation was obtained from 500
samples of a set of uniform random numbers whose size is equal to
the number of spikes recorded at that stimulus intensity.
A more rigorous evaluation of fiber independence is a
recurrence-time test. (See, for example, Johnson and Kiang.) By
using a bin size of 0.5 ms, useful recurrence-time histograms were
assembled from two 2-second spike trains of the ANF model
simulation. FIG. 8A shows a 50 ms sample of spike activity from two
"units" (i.e., two simulated neurons). FIG. 8B shows an ISI
histogram from an eight second run of "unit" b. FIG. 8C shows a
forward recurrence-time histogram of "unit" b to "unit" a, and a
theoretical recurrence-time from "unit" b assuming that "units" a
and b are independent. The theoretical forward recurrence-time
curve is flat during the refractory period. Theoretical limits are
shown at .rho.<0.0124 (2.5 standard deviations). FIG. 8D shows
residuals calculated by subtracting the curves in FIG. 8C. Thus,
the ANF model demonstrated pseudospontaneous activity caused by
high rate pulse train stimulation.
FIG. 9 shows increases in pulse rates above 1016/s. In particular,
FIG. 9 shows magnitudes of electrically-evoked compound action
potentials (EAPs) produced with stimulation of human implant
subject intracochlear electrodes with identical pulses presented at
varied rates. The magnitudes were normalized to the magnitude of
the EAP following the first pulse. The pulse amplitude was 375 mA
and the pulse duration was 33 .mu.s/phase. Stimulations were
applied using one electrode (i.e., electrode 3) and recordings were
made with an adjacent electrode (i.e., electrode 4). Body potential
was measured with a reference electrode at the wrist. Methods for
generating EAP responses are known in the art. EAP responses in
FIG. 9 were determined using a subtraction technique to remove the
influence of all prior stimuli and the corresponding responses from
the response to a final pulse in a train. In other words, the
response to the Nth pulse for each condition was determined by
subtracting a record for an N-1 pulse train from a record for an N
pulse train, which leaves only the response to the Nth pulse.
Without such subtraction, prior EAPs would overlap because the
approximately 1 ms duration of an EAP waveform was greater than the
interval between sequential pulses and EAPs for pulse rates greater
than about 1000/s.
As shown in FIG. 9, the magnitudes of EAPs were produced with
stimulation of an intracochlear implant with identical pulses
respectively presented at the rates of 1016, 2033, 3049 and 4065/s.
Increases in pulse rate to 3049/s or higher produced uniform
magnitudes of sequential EAPs after the first millisecond of
stimulation. A large EAP is elicited by the first pulse, followed
by a transient depression in excitability, and then by uniform
response amplitudes 902. The transient depression in excitability
may be caused by the refractory period. The uniform response result
corresponds with the ANF model simulation results shown in FIG. 4.
The constant response amplitude after 1 ms in FIG. 9 is likely
caused by a different, possibly equal sized, pool of fibers
responding to each pulse. The constant response amplitude is likely
the EAP manifestation of stochastic independence at the single-unit
level demonstrated by the ANF model simulations.
If pseudospontaneous activity can be created by a driving signal
according to the first preferred embodiment such as high-rate
constant pulse train, the auditory nerve can be desynchronized
using such a stimulus. Desynchronization of the auditory nerve has
various benefits. For example, desynchronization can improve
temporal representation of a modulated stimulus. Further,
desynchronized auditory nerve responses are a closer match to
responses detected in the normal, synaptically driven nerve.
FIG. 10 shows desynchronization of an auditory nerve. As shown in
FIG. 10 a high rate conditioner (e.g., driving signal) was combined
with a low rate stimulus. The conditioner starts at time zero and
includes identical pulses presented at the rate of 5039/s. The
stimulus starts 29 ms after the onset of the conditioner and
includes identical pulses presented at the rate of 1016/s. The
amplitude of the stimulus pulses in FIG. 10 were 375 .mu.A, as in
FIG. 8. The amplitude of the conditioner was varied between a 100
.mu.A conditioner and a 375 .mu.A conditioner. As shown, EAP
magnitudes for the stimulus are normalized to the magnitude of the
EAP following the first pulse of the stimulus for the "no
conditioner" case. EAPs following the pulses of the stimulus were
derived using the subtraction technique similar to FIG. 9. In other
words, recording conditions were identical for FIGS. 9-10. FIG. 11
shows an exemplary waveform of a conditioner 1102 and a stimulus
1104.
Thus, increases in the conditioner amplitude from about 200 .mu.A
to 300 .mu.A produced substantial changes in the pattern of
responses to the stimulus. In particular, responses become more
uniform with increases in conditioner amplitude over this range.
Further, increases in conditioner or driving signal amplitude
produced decrements in the magnitude of the EAPs, but do not change
the uniform pattern of responses across pulses. The neural
representation of the deterministic stimulus is much improved by
the addition of the conditioner for conditioner amplitudes at and
above about 250 .mu.A. High levels of responses to the stimuli are
maintained with conditioner amplitudes as high as 325 .mu.A. Thus,
relatively large numbers of neurons can be available for
representation of the deterministic stimulus over this range.
As described above, driving a population of simulated auditory
nerve fibers with high rate pulses according to the first preferred
embodiment produces independent spike trains in each simulated
fiber after about 20 ms. This pseudospontaneous activity is
consistent with a renewal process and yields statistical data
comparable to true spontaneous activity within computational
limitations. However, the first preferred embodiment of the
invention is not intended to be limited to the above. For example,
broadband additive noise (e.g., because of rapid signal amplitude
transitions) could also evoke pseudospontaneous activity similar to
that induced by the high rate pulse train. Any signal that results
in pseudospontaneous activity that meets the same tests of
independence as true spontaneous activity can be used as the
driving signal.
As alluded to hereinabove, in a normal cochlea the inner hair
cell-spiral ganglion is inherently "noisy" (i.e., there is a high
background of activity in the absence of sound) resulting in
spontaneous activity in the auditory nerve. Further, sound produces
a slowly progressive response within and across fiber
synchronization as intensity is increased. Deficiencies in the
perception of sounds by a patient having a defective cochlea fitted
with a cochlear implant and receives electrical stimulation using a
conventional or prior art speech processing system may be
explained, at least in part, by the absence of spontaneous neural
activation of the auditory nerve.
In a second preferred embodiment of an inner ear implant according
to the present invention, a conditioner or driving signal and a
speech signal are provided by the implant. Thus, the induction of
pseudospontaneous activation of the auditory nerve by the delivery
of a high rate pulse train can be integrated with a speech
processing strategy such that the high rate pulse train is
superimposed on the speech signal, or added to the speech signal,
etc. Desynchronizing the auditory nerve by means of a high rate
pulse train, for example, improves the nerve's representation of
the temporal details of a speech processor signal, while
characteristics of normal hearing such as wide dynamic range both
within and across auditory neurons can be restored. These features
compensate for deficiencies noted herein above for cochlear
implants/speech processing systems of the related art.
As shown in FIG. 12, the second preferred embodiment includes an
inner ear stimulation system 1200 that directly electrically
stimulates the auditory nerve (not shown). The inner ear
stimulation system 1200 can include two components: (1) a wearable
or external system, and (2) an implantable system.
An external system 1202 includes a signal generator 1210, a
microphone 1204 and a pseudospontaneous signal generator 1206. The
signal generator 1210 can include a battery, or an additional
equivalent power source 1214, and further includes electronic
circuitry, typically including a controller 1208 that controls the
signal generator 1210 to produce prescribed electrical signals
1216. The signal generator 1210 not only produces the electrical
signals 1216 to electrically simulate speech but also to generate
pseudospontaneous activity in the auditory nerve.
The signal generator 1210 combines a driving signal from the
pseudospontaneous signal generator 1206 and a signal that
represents sound received from the microphone 1204 or the like.
Preferably, the signal generator adds the signals from the
pseudospontaneous signal generator 1206 and the microphone 1204.
The pseudospontaneous signal generator 1206 can produce a driving
signal in accordance with the first preferred embodiment. The
pseudospontaneous signal generator 1206 can be a signal generator.
However, any device that produces a waveform that generates
pseudospontaneous activity can be used. That is, any device that
produces a pseudospontaneous driving signal. For example, an
application program operating on a special purpose computer or
microcomputer combined with an A/D converter, a LC resonating
circuit, firmware or the like can be used, depending on the exact
form of the pseudospontaneous driving signal.
The inner ear stimulation system 1200 provides an improved hearing
response to the signal from the microphone 1204 that represents
sound. The signal generator 1210 can further vary parameters such
as the frequency, amplitude and pulse width of the electrical
signals 1216. The external system 1202 can be coupled to a head
piece 1212. For example, the head piece 1212 can be an ear piece
worn like a hearing aid. Alternatively, the external system 1202
can be a separate unit.
The controller 1208 is preferably implemented on a microprocessor.
However, the controller 1208 can also be implemented on a special
purpose computer, microcontroller and peripheral integrated circuit
elements, an ASIC or other integrated circuit, a hardwired
electronic or logic circuit such as a discrete element circuit, a
programmable logic device such as a PLD, PLA, FGPA or PAL, or the
like. In general, any device on which a finite state machine
capable of controlling a signal generator and implementing the
flowchart shown in FIG. 14 can be used to implement the controller
1208.
As shown in FIG. 12, an implantable system 1220 of the inner ear
stimulation system 1200 can include a stimulator unit 1222 directly
coupled to the auditory nerve via implementation in the cochlea
(not shown). For example, the stimulator unit 1222 can include an
electrode array 1224 or the like implanted into the cochlea of a
patient. The electrode array 1224 can be a single electrode or
multiple electrodes that stimulate different sites at discrete
locations within along the cochlea to evoke nerve activity normally
originating from the respective discrete locations. In addition,
the implantable system 1220 can be directly or indirectly coupled
to the external system 1202.
If indirectly coupled to the external system 1202, the stimulator
1222 can include a receiver 1226. The receiver 1226 can receive
information and power from corresponding elements in the external
system 1202 through a receiving coil (not shown) attached to the
receiver 1226. The power, and data as to which electrode to
stimulate, and with what intensity, can be transmitted across the
skin using an inductive link from the external signal generator
1210. For example, the receiver 1226 can then provide the signals
1216 to the electrode array 1224. Alternatively, the stimulation
unit 1222 can be directly coupled to the external system 1202 via a
conductive medium or the like.
Upon installation and periodically thereafter, the patient's
hearing based on the electrical signals 1216 can be subsequently
monitored or tested. The results of these tests could be used to
modify the electrical signal 1216 or select from a plurality of
pseudospontaneous driving signals using a selection unit 1218.
The stimulation unit 1222 can operate in multiple modes such as,
the "multipolar" or "common ground" stimulation, and "bipolar"
stimulation modes. However, the present invention is not intended
to be limited to the above. For example, a multipolar or
distributed ground system could be used, wherein all other
electrodes do not act as a distributed ground, and any electrode
could be selected at any time to be a current source, a current
sink, or to be inactive during either stimulation phase with
suitable modification of the receiver-stimulator. Thus, there can
be flexibility in the choice of a stimulation strategy by the
stimulation unit 1222 to provide the electrical signals 1216 to the
auditory nerve. However, the specific method or combination of
electrodes in the electrode array 1224 used to apply the driving
signal must result in the pseudospontaneous activity being
generated and a signal representing sound being provided. The
present invention is not intended to be limited to a specific
design of the electrode array 1224, but rather a number of
alternative electrode designs as have been described in the prior
art could be used.
FIGS. 13A and 13B show exemplary implementations of the inner ear
stimulation system 1200. The pseudospontaneous signal generator
1206 driving signal is combined with a signal from a speech
processor 1310. The rate of the speech processing signal is less
than the rate of the driving signal. A number of alternative speech
processor designs as have been described in the prior art can be
used. As shown in FIG. 13A, the speech signal is preferably added
to the driving signal in a combining circuit 1320. However, the
signals can be superimposed or modulated together in the combining
circuit 1320. A combined signal is then transmitted to a electrode
1330 of an electrode array (not shown). FIG. 13B shows the speech
signal and the driving signal being delivered separately to the
auditory nerve (not shown) using the inner ear stimulation system
1200. Electrodes 1330 and 1330 represent individual electrodes or
electrode subsets in an electrode array.
A third preferred embodiment of a method for speech processing
using pseudospontaneous stimulation according to the present
invention will now be described. As shown in FIG. 14, the process
starts in step S1400. From step S1400, control continues to step
S1410. In step S1410, a signal representing sound is received. For
example, the signal could be individually generated or generated
from a combination of a microphone, prerecordings, or input from a
plurality of sources (e.g., a television, etc.). From step S1410,
control continues to step S1420.
In step S1420, a pseudospontaneous driving signal is generated. For
example, a driving signal according to the first preferred
embodiment can be generated or selected in step S1420. From step
S1420, control continues to step S1430. In step S1430, the driving
signal and the first signal are combined. From step S1430, control
continues to step S1440.
In step S1440, the combined signal is applied to the auditory
nerve. For example, an inner ear implant according to the second
preferred embodiment can be used to implement a method according to
the third preferred embodiment. Because of the pseudospontaneous
activity generated in the auditory nerve by the driving signal, the
response of the auditory nerve to the signal representing sound is
improved. Even with a broad range of electrical thresholds in the
auditory nerve (approximately 12 dB), near physiologic rates may be
maintained across most of the auditory nerve with multiple
electrodes. From step S1440, control continues to step S1450 where
the process is completed. The method according to the third
preferred embodiment can optionally include a feed-back test loop
to modify or merely select one of a plurality of selectable
pseudospontaneous driving signals based on a subset of parameters
specifically designed and determined for an individual patient.
As described above, the preferred embodiments according to the
present invention have various advantages. The preferred
embodiments generate stochastically independent or
pseudospontaneous neural activity, for example, in an auditory
nerve to improve a speech processing apparatus and method. Further,
the stimulus that evokes pseudospontaneous activity should not be
perceptible over the long term as long as the rate is physiologic.
According to the preferred embodiments, a conditioner signal and a
data signal can be combined and the combined signal is provided to
a neural system to improve the response of the neural system to the
data signal. For example, an inner ear implant such as a cochlear
implant, according to the preferred embodiments generates
pseudospontaneous activity in an auditory nerve, which improves a
response to signals representing sound.
The foregoing embodiments are merely exemplary and are not to be
construed as limiting the present invention. The present teaching
can be readily applied to other types of apparatuses. The
description of the present invention is intended to be
illustrative, and not to limit the scope of the claims. Many
alternatives, modifications, and variations will be apparent to
those skilled in the art.
* * * * *